In this project I have generated images similar to the popular MNIST dataset using the technique of GAN(Generative Adverserial Network) in PyTorch. As an output a .avi file is generated which shows the transition of images getting better as and when the model is trained for more number of epochs.
The output looks something like this: (Play the video to check the output)